A Low Complexity Near Optimal Signal Detection for Large Scale MIMO SC-FDMA Uplink System

This article proposes an improved Newton algorithm as a low complexity signal detection scheme for linear receiver in large scale multiple- input multiple- output (LS-MIMO) single carrier frequency division multiple access (SC-FDMA) uplink system, where a large number of antennas are set up at the base station and active users are with a single antenna system. Data detection for uplink SC-FDMA system is one of the specific challenges due to the significant rise in the dimension of antennas and number of subcarriers. Especially for symbol detection process, LS-MIMO SC-FDMA system with linear detector requires to perform a large matrix inverse computation. Even though linear detectors such as zero forcing (ZF) and minimum mean square error (MMSE) can achieve near-optimal detection performance, they still introduce high computational complexity and obliviously involve in the computation of matrix inversion. Therefore, a design of complexity reduction algorithm based near-optimal detector for LS-MIMO SC-FDMA system attains research interest. The improved Newton algorithm is employed to obtain linear detection solution which iteratively performs matrix free-inversion operation. The new algorithm performs matrix–matrix multiplication into matrix–vector multiplication, which substantially reduces receiver detection complexity. The efficacy of the proposed method is investigated at 16-QAM. Both ZF and MMSE criteria are proposed and compared through simulations. Simulation results illustrate that the proposed scheme outperforms the conventional detection schemes and exhibits near-optimal performance with a small number of iterations. Further, bit-error-rate performance is closer to classical linear detector with affordable computational complexity.